What it is:Kidney damage or reduced function that has stuck around for 3 months or longer. Not a single disease, a final common pathway that diabetes, hypertension, glomerulonephritis, and a dozen other insults all funnel into.
The core problem:Nephrons are lost faster than they regenerate (they don't regenerate). The nephrons that survive compensate by hyperfiltering, and that compensation is exactly what destroys them next. It's a slow-motion, self-feeding decline, and the disease itself doesn't hurt until it's very advanced.
What you do about it:Stage it correctly (Cause-GFR-Albuminuria), then stack the drugs that lower pressure inside the glomerulus: ACEi/ARB, SGLT2 inhibitors, and finerenone. Everything else on this page (anemia, bone/mineral disease, hyperkalemia) is managing the downstream damage once kidney function has already fallen.
Screening beats symptom-watching. CKD is silent for years, so you find it by checking eGFR and urine albumin in people with risk factors (diabetes, hypertension, CVD, family history, older age), not by waiting for them to feel sick. By the time someone feels uremic, they've usually lost most of their nephron mass already.
Two separate lab values (eGFR and urine albumin-to-creatinine ratio, uACR) get combined into one risk grid. Both need to be present for 3 months or morebefore you can call it chronic. That duration requirement is what separates CKD from AKI, which is the same abnormal labs but present for days to weeks and often reversible.
CKD = abnormality of kidney structure or function present >3 months, where the abnormality is EITHERdecreased GFR (<60 mL/min/1.73m²) ORa marker of kidney damage: albuminuria (uACR ≥30 mg/g), abnormal urine sediment, persistent hematuria, electrolyte abnormalities from a tubular disorder, structural abnormalities on imaging, histologic abnormalities on biopsy, or a history of kidney transplant.
| Stage | eGFR (mL/min/1.73m²) |
|---|---|
| G1 | ≥90 |
| G2 | 60–89 |
| G3a | 45–59 |
| G3b | 30–44 |
| G4 | 15–29 |
| G5 | <15 or on dialysis |
| Category | uACR | Terms |
|---|---|---|
| A1 | <30 mg/g | Normal to mildly increased |
| A2 | 30–300 mg/g | Moderately increased |
| A3 | >300 mg/g | Severely increased |
G1 and G2 alone do not equal CKD.If eGFR is ≥60 and there's no albuminuria and no other structural marker, the patient does not meet criteria, no matter how many risk factors they have. You need the ≥3-month duration plus at least one qualifying abnormality. A 45-year-old with eGFR 82 and confirmed uACR 180 mg/g has G2A2 CKDeven though their GFR looks fine, because the albuminuria is the qualifying damage marker.
Combine G and A into a single risk grid (the KDIGO "heat map") and you get four risk tiers for progression, cardiovascular events, and mortality: low risk(G1–G2, A1, and only if another damage marker is present), moderately increased(G1–G2/A2, G3a/A1), high risk(G1–G2/A3, G3a/A2, G3b/A1), and very high risk(G3a/A3, G3b/A2–A3, and everything in G4 and G5 regardless of albuminuria). Risk climbs moving right (more albuminuria) faster than it climbs moving down (lower GFR), which is exactly why albuminuria gets its own axis instead of being an afterthought.
The "C" in CGA is the underlying etiology: diabetic kidney disease, hypertensive nephrosclerosis, glomerulonephritis, polycystic kidney disease, obstructive uropathy, and so on. Diabetes and hypertension together account for the large majority of new ESKD in the US. Cause matters because it changes prognosis and sometimes changes therapy (immunosuppression for glomerulonephritis, for example), but for pharmacotherapy purposes G and A drive almost every decision on this page.
Nearly every progression-slowing drug in CKD targets the same mechanism: pressure inside the glomerulus. Understand that and the whole drug list stops being a memorization exercise.
Whatever the initial injury (diabetes, hypertension, glomerulonephritis, whatever), some nephrons die. The ones left behind don't just sit there doing proportionally less work. They hyperfilter: the afferent arteriole dilates and the efferent arteriole constricts, which raises glomerular capillary pressure and glomerular blood flow so the surviving nephrons can cover for the ones that were lost.
Hyperfiltration feels adaptive in the short term, more filtration per nephron keeps total GFR from crashing, but the sustained high pressure is itself toxic. It causes glomerular hypertrophy, mesangial expansion and injury, podocyte (epithelial) foot process detachment, and endothelial injury. That damage produces proteinuria, and the resulting glomerulosclerosis and arteriosclerosis kill more nephrons, which forces the remaining ones to hyperfilter even harder. It's a self-amplifying loop, and every nephroprotective drug you'll learn works by interrupting it, not by fixing the original insult.
| Drug class | What it does to the glomerulus |
|---|---|
| ACEi / ARB | Relaxes the efferent arteriole preferentially, dropping intraglomerular pressure. This is why they cause an early eGFR dip, less pressure means less filtration, and why they reduce proteinuria as a class effect. |
| SGLT2 inhibitors | Block glucose/sodium reabsorption in the proximal tubule, which increases sodium delivery to the macula densa. That restores tubuloglomerular feedback, constricting the afferent arteriole and lowering glomerular pressure from the other side of the equation. |
| Finerenone (nsMRA) | Doesn't primarily change hemodynamics. Blocks mineralocorticoid receptor-driven fibrosis, inflammation, and oxidative stress in the kidney and heart. |
| Non-DHP CCBs | Dilate the afferent arteriole (opposite of DHP CCBs), which is why they're antiproteinuric while dihydropyridines like amlodipine generally aren't when used alone. |
Starting an ACEi/ARB, SGLT2i, nsMRA, or GLP-1RA drops eGFR a little right after initiation. A decrease under 30% is expected and reversible, it reflects lower pressure across fewer, less-stressed nephrons, not injury. Long-term, that lower pressure is exactly what slows progression. Don't stop the drug for an isolated small creatinine bump; confirm it isn't from volume depletion, then keep going.
CKD is generally asymptomatic until it's advanced. Symptoms, when they show up, are nonspecific and easy to blame on something else.
| Stage | What you'll see |
|---|---|
| Early (G1–G3) | Usually nothing. May have hypertension or mild edema. Found on routine labs or diabetes screening, not because the patient complained. |
| Advanced (G4–G5) | Fatigue, anorexia, nausea, pruritus, cognitive slowing, sleep disturbance, edema/volume overload, and eventually the full uremic picture as filtration failure lets toxins accumulate. |
Because symptoms show up so late, risk-factor-based screening is far more useful than symptom-based detection. That's the whole rationale for annual CKD screening in diabetes: you're trying to catch G2A2 before it becomes G4A3, not waiting for someone to feel bad.
A frail, low-muscle-mass patient can have a serum creatinine of 0.7 mg/dL and a calculated eGFR of 88, and still have real, worse kidney disease, because they don't generate enough creatinine for the number to be trustworthy. Cystatin Cis produced by essentially all nucleated cells at a constant rate, independent of muscle mass, so it's the better biomarker to reach for when body composition makes creatinine suspect.
Screen type 2 diabetics annually starting at diagnosis(they may have had silent hyperglycemia for years before diagnosis). Screen type 1 diabetics starting 5 years after diagnosis, since the onset is usually known and the kidney damage takes time to develop. Screening means SCr/eGFR plus uACR together, not one or the other.
A patient with diabetes, hypertension, eGFR 95, and confirmed uACR 180 mg/g has real CKD-related risk right now (that's still G2A2 territory functionally, even with a "normal" eGFR). Elevated albuminuria independently predicts progression and cardiovascular events. Treat the albuminuria and the hyperfiltration driving it; don't defer therapy until creatinine finally rises, by then you've lost the window where intervention does the most good.
Four drug classes now carry evidence for slowing CKD progression independent of, or on top of, blood pressure control. They're not interchangeable, they're additive, and unlike heart failure's four pillars, they don't all start on day one for every patient, ACEi/ARB and SGLT2i are the backbone and the others layer on based on albuminuria, diabetes status, and tolerability.
Any A2 or greater albuminuria (uACR >30 mg/g). Antiproteinuric effect is a class effect.
Slows progression in diabetic andnondiabetic proteinuric CKD. Reduces dialysis/transplant need and mortality.
Studied in T2D-related CKD. Option if SGLT2i not tolerated; not studied combined with SGLT2i + ACEi/ARB together.
Newest addition to the stack. Slows CKD progression and lowers CV risk on top of the above.
Third-line antiproteinuric option when ACEi/ARB andSGLT2i are both contraindicated or not tolerated: non-dihydropyridine CCBs(diltiazem, verapamil). They're not recommended if the patient is already on a beta blocker, for the usual additive bradycardia/AV-block concern.
| Class | Slows CKD | ↓ Glucose | ↓ CV risk | ↓ HF hosp. | ↓ BP | ↓ Weight |
|---|---|---|---|---|---|---|
| ACEi / ARB | ✓ | ✓ | ✓ | ✓ | ||
| SGLT2 inhibitors | ✓ | ✓ | ✓ | ✓ | modest | |
| Finerenone (nsMRA) | ✓ | ✓ | ✓ | modest | ||
| GLP-1 (±GIP) RA | ✓ | ✓ | ✓ | ✓ | ✓ | |
| Statins | ✓ | |||||
| Antiplatelets | ✓ |
General class effects, not FDA labeling for every individual agent within a class.
Once a patient is CKD G1–G4 with A2 or A3 albuminuria, start an ACEi or ARB and target SBP <120 mmHg. Check BP, SCr, and potassium 2–4 weeksafter starting or increasing the dose. Keep the drug going unless SCr rises more than 30% from baseline. If BP is still above goal: increase the ACEi/ARB dose and/or add a thiazide (switch to a loop diuretic if the patient has edema) or a CCB. Still not at goal: add an MRA (especially useful if the patient is hypokalemic), a beta blocker, or an alpha blocker. Still not at goal: add clonidine, hydralazine, or minoxidil.
If the patient is CKD G1–G4 but A1(no significant albuminuria), there's no special low BP target tied to CKD, just treat hypertension per general guidelines: ACEi, ARB, CCB, and thiazides are all reasonable first-line choices.
Contraindicated:pregnancy (confirm reliable contraception in anyone of childbearing potential), bilateral renal artery stenosis, prior angioedema with an ACEi or ARB. Hold temporarilyduring vomiting, diarrhea, or any intravascular volume depletion, that combination sets up pre-renal AKI. Never combinean ACEi with an ARB, or either with a direct renin inhibitor.
Managing diabetes in a CKD patient is really about reducing proteinuria and hitting BP and A1c targets simultaneously, the three goals reinforce each other. This is where the nephroprotective stack and glycemic management overlap almost completely (SGLT2i and GLP-1RA do double duty).
These are the drugs with real numbers worth knowing cold: erythropoiesis-stimulating agents, IV iron, phosphate binders, vitamin D agents, and calcimimetics.
| Drug | Brand(s) | Starting dose | Route | Half-life |
|---|---|---|---|---|
| Epoetin alfa | Epogen, Procrit | 50–100 units/kg 3x/week (adult) | IV or SubQ | 8.5h IV / 24h SubQ |
| Epoetin alfa-epbx | Retacrit | Biosimilar, same dosing as epoetin alfa | ||
| Darbepoetin alfa | Aranesp | 0.45 mcg/kg q4wk (ND-CKD) or weekly (ESKD) | IV or SubQ | 25h IV / 48h SubQ |
| Methoxy PEG-epoetin beta | Mircera | 0.6 mcg/kg q2wk, then double the dose and give monthly once Hb stabilizes | IV or SubQ | 134h IV / 139h SubQ |
SubQ dosing has a slower absorption phase, which stretches the effective half-life and lets patients maintain target Hb on 15–30% less drugthan the IV route. Darbepoetin and methoxy PEG-epoetin beta have long enough half-lives to allow much less frequent dosing than epoetin alfa, that's their whole selling point.
| Drug | Brand | Notable dosing |
|---|---|---|
| Ferric carboxymaltose | Injectafer | 750 mg x2 doses ≥7 days apart (≥50 kg); max 1500 mg/course |
| Ferric derisomaltose | Monoferric | 1000 mg single dose (≥50 kg) |
| Ferric pyrophosphate citrate | Triferic / Triferic AVNU | Added to dialysate, continuous iron delivery during HD |
| Ferumoxytol | Feraheme | 510 mg, then a second 510 mg dose 3–8 days later |
| Iron dextran | INFeD | 100 mg over 2 min; 25 mg test dose required first |
| Iron sucrose | Venofer | 100 mg over 2–5 min per HD session (adult) |
| Sodium ferric gluconate | Ferrlecit | 125 mg over 10 min (HD patients on an ESA) |
Iron dextran carries the highest anaphylaxis riskof any IV iron product (hence the mandatory test dose), and iron sucrose carries the lowest. Monitor patients for at least 30 minutes after any infusion (KDIGO suggests 60 minutes, more conservative for iron dextran at a 1B recommendation vs 2C for non-dextran products). Long-term IV iron use also risks iron overload, hepatic, pancreatic, and cardiac dysfunction, so keep ferritin and TSat within target rather than repeatedly dosing on autopilot.
| Category | Drug | Brand | Starting dose | Key point |
|---|---|---|---|---|
| Calcium-based (first-line) | ||||
| Calcium acetate | PhosLo, Phoslyra | 1334 mg TID with meals | ~45 mg phosphorus bound per 1 g | |
| Calcium carbonate | Tums, Os-Cal, Caltrate | 0.5–1 g elemental calcium TID with meals | ~39 mg phosphorus bound per 1 g | |
| Iron-based | ||||
| Ferric citrate | Auryxia | 420 mg TID with meals | May raise serum iron/ferritin/TSat; dark stools | |
| Sucroferric oxyhydroxide | Velphoro | 500 mg TID with meals | Dark stools | |
| Resin (nonabsorbed) | ||||
| Sevelamer carbonate | Renvela | 800–1600 mg TID with meals | Also ↓LDL, ↑HDL; lower metabolic acidosis risk than the HCl salt | |
| Sevelamer hydrochloride | Renagel | 800–1600 mg TID with meals | Risk of metabolic acidosis | |
| Other | ||||
| Lanthanum carbonate | Fosrenol | 1500 mg/day divided, with meals | Lower hypercalcemia risk; long-term GI accumulation of unclear significance | |
| Aluminum hydroxide not preferred | AlternaGel | 300–600 mg TID with meals | Aluminum toxicity risk; short-term use only (≤4 weeks) | |
| Category | Drug | Brand | Initial dose | Note |
|---|---|---|---|---|
| Nutritional | ||||
| Ergocalciferol (D2) | Drisdol | Varies by 25(OH)D level | Daily or weekly/monthly for high doses | |
| Cholecalciferol (D3) | Generic | Varies | Weekly/monthly for higher doses | |
| Calcifediol | Rayaldee | 30 mcg daily | Increase after 3 months if PTH still elevated | |
| Active vitamin D / analogs | ||||
| Calcitriol (PO) | Rocaltrol | 0.25 mcg daily | ↑ by 0.25 mcg/day q4–8wk | |
| Calcitriol (IV) | Calcijex | 1–2 mcg TIW | ↑ by 0.5–1 mcg q2–4wk | |
| Doxercalciferol | Hectorol | 1 mcg daily (ND-CKD); 10 mcg TIW (ESKD, PO) | Prodrug, needs hepatic activation | |
| Paricalcitol | Zemplar | 1 mcg daily (ND-CKD, PTH ≤500) | Less hypercalcemia/hyperphosphatemia than calcitriol | |
KDIGO does notrecommend routine active vitamin D (calcitriol or an analog) in nondialysis CKD. Reserve it for CKD stages 4–5. Confirm calcium and phosphorus are already within normal range before starting, and adjust the dose every 2–4 weeks based on PTH trend, not on a fixed schedule.
| Drug | Route | Starting dose | Max / titration |
|---|---|---|---|
| Cinacalcet | PO | 30 mg daily | Titrate q2–4wk to max 180 mg daily |
| Etelcalcetide | IV, at end of HD | 5 mg TIW | Titrate per Ca/PTH response |
Both work by sensitizing the calcium-sensing receptor, which shuts off PTH secretion. Because they also lower serum calcium, don't start either one if calcium is already below the lower limit of normal.Check calcium and phosphorus 1 week after starting, and PTH within 1–4 weeks or per the dialysis unit's protocol.
Common ADE unique to ACEi: cough(bradykinin accumulation). Rare-but-serious for both: angioedema. Absolute contraindications for both: history of angioedema (hereditary, idiopathic, or drug-related) and pregnancy (Black Box Warning). Use cautiously with dehydration/volume depletion and in breastfeeding.
Common ADEs: genitourinary infections, hypoglycemia (mostly when combined with insulin/sulfonylureas). Rare but serious: euglycemic DKAand Fournier's gangrene(necrotizing infection of the perineum). Contraindicated on dialysis, since there's no filtrate left for the drug to act on. Precautions: volume depletion, pregnancy (2nd/3rd trimester), breastfeeding.
Bexagliflozin and ertugliflozinhave weaker or absent CKD-outcome data compared to empagliflozin, dapagliflozin, and canagliflozin, worth knowing if a question is testing which agent has trial support.
Finerenone and etelcalcetide's cousin-in-name-only, don't confuse them, finerenone reduces PTH secretion nowhere, it's a mineralocorticoid receptor antagonist that blocks aldosterone's fibrotic and inflammatory effects on kidney and heart tissue. Common ADEs: hyperkalemia(the dose-limiting toxicity, watch it closely) and, for steroidal MRAs specifically, gynecomastia with spironolactone. Contraindications: hyperkalemia, adrenal insufficiency (Addison's disease), and for finerenone specifically, strong CYP3A4 inhibitors(it's a substrate). Use cautiously in pregnancy, breastfeeding, and low eGFR.
Finerenone has the strongest trial evidence for CKD specifically. Spironolactone is used when cost is the deciding factor. Eplerenone is the swap if a patient is spironolactone-intolerant, though that's less common in CKD than in heart failure. All nsMRAs cause hyperkalemia more readily than ACEi/ARB alone, which is why potassium monitoring matters even more once you layer one on.
The primary driver is a relative deficiency of erythropoietinproduction by the failing kidney. Contributing factors stack on top: reduced GI iron absorption, chronic inflammation, frequent blood draws, blood loss during hemodialysis, increased iron demand once an ESA is started, shortened RBC lifespan, accumulation of uremic toxins and inflammatory cytokines, and B12/folate deficiency.
| ND-CKD | ESRD | |
|---|---|---|
| Start an ESA if | Hb <10 g/dL, after weighing rate of Hb fall, prior iron response, transfusion risk, and symptoms | Do not start if Hb ≥10; start when Hb is 9–10 to avoid a drop below 9 |
| Hb ceiling | Never intentionally push Hb above 13 g/dL; don't use an ESA to maintain Hb above 11–11.5 g/dL | |
| Start iron if | TSat ≤30% and ferritin ≤500 ng/mL | |
All ESAs carry a warning that targeting Hb above 11 g/dLincreases the risk of death, serious cardiovascular events, and stroke. This isn't a soft suggestion, it's the reason ESA dosing is titrated to a range, not to "normal."
ESA hyporesponsiveness(needing escalating doses without benefit): think iron deficiency, acute illness, inflammation/infection, ongoing bleeding, aluminum toxicity, malnutrition, hyperparathyroidism, or malignancy/chemo. Never escalate an ESA dose beyond double the initial weight-based dose without working through this list first. If Hb rises too fast, cut the ESA dose by 25%; if it's not rising enough after 4 weeks, increase by 25% (after ruling out the causes above). The most common ESA adverse effect is hypertension, tied to how fast Hb climbs.
As GFR falls, phosphate excretion falls with it, causing hyperphosphatemia. Rising phosphate drives serum calcium down (they're reciprocally linked), and hypocalcemia stimulates PTH secretion. That secondary hyperparathyroidism increases bone resorption to free up calcium, at the cost of bone structural integrity. Reduced kidney mass also means less calcitriol production, compounding the hypocalcemia and PTH rise from a second angle.
| Parameter | Target |
|---|---|
| Calcium | Avoid hypercalcemia at every CKD stage (higher calcium tracks with mortality and CV events) |
| Phosphorus | "Toward the normal range" at every stage (both high and low phosphate track with mortality) |
| PTH | Avoid progressively rising or persistently above-normal levels; in ESKD, target 2–9x the upper normal limit |
Monitoring frequency for all three intensifies as CKD advances, roughly every 6–12 months at G3, down to every 1–3 months at G5/ESKD. Nonpharmacologic tools come first: dietary phosphorus restriction, dialysis itself, and parathyroidectomy for refractory hyperparathyroidism. Pharmacologically, phosphate binders come first, active vitamin D is reserved for stage 4–5 and not used routinely, and calcimimetics target PTH directly when binders and vitamin D aren't enough.
CKD patients get hyperkalemic for two reasons stacked together: reduced renal potassium excretion, and the RAAS-blocking drugs (ACEi/ARB, nsMRA) you're deliberately using because they protect the kidney. The instinct to just stop the offending drug at the first elevated potassium is usually the wrong move, since that drug is often the single most important thing slowing progression.
A patient on losartan for CKD/heart failure with no fluid overload has a potassium of 5.4 mEq/L, and a diet history reveals frequent salt-substitute use. The fix is eliminating the salt substitute, not reducing losartan and definitely not adding spironolactone (which would make the hyperkalemia worse, not better).
CKD is an independent cardiovascular risk multiplier, and CV disease is the leading cause of death in CKD, well before most patients ever reach dialysis. Traditional risk factors (diabetes, dyslipidemia, hypertension, smoking, obesity) combine with CKD-specific, nontraditional ones: proteinuria, anemia, chronic inflammation, and abnormal calcium/phosphate handling driving vascular calcification and oxidative stress.
Aspirinis recommended for secondary prevention in essentially all CKD patients based on observational mortality benefit. Beyond aspirin, non-dialysis CKD patients have several options to delay progression while simultaneously lowering CV risk: SGLT2 inhibitors, GLP-1 receptor agonists, and nsMRAs, the same nephroprotective stack from the treatment section, doing double duty.
| # | Rule |
|---|---|
| 1 | Get a fasting lipid panel in every adult newly diagnosed with CKD. |
| 2 | Start a statin in adults 18–49 years with CKD (not on dialysis/transplant) who have known coronary disease, diabetes, prior ischemic stroke, or an estimated 10-year coronary death/MI risk >10%. |
| 3 | Start a statin, or statin/ezetimibe, in adults >50 years with eGFR <60 (not on dialysis/transplant). |
| 4 | Do notinitiate a statin or statin/ezetimibe in dialysis-dependent CKD, but continue it if the patient was already taking it when dialysis started. |
Unlike general dyslipidemia management, KDIGO does not recommend titrating statin dose to a specific LDL target in CKD. You start a standard dose based on the risk category above and you're done, no periodic lipid-panel-driven up-titration. This trips people up because it breaks the "treat to target" habit from every other lipid guideline.
| Parameter | When | Watching for |
|---|---|---|
| BP, SCr, K+ | 2–4 weeks after starting or increasing an ACEi/ARB | Renal decline (hold if SCr ↑>30%), hyperkalemia |
| eGFR, uACR | Periodically, frequency scales with risk category | Progression, need to escalate the nephroprotective stack |
| Hb, TSat, ferritin | Monthly (Hb) once on an ESA; TSat/ferritin at least every 3 months | Overshoot/undershoot of Hb target, iron deficiency |
| Calcium, phosphorus | Every 1–3 months (G5/ESKD) to every 6–12 months (G3), and 1 week after starting a calcimimetic | Hypercalcemia, phosphorus drifting from normal |
| Intact PTH | Every 3–12 months depending on stage; 1–4 weeks after starting a calcimimetic | Rising or persistently elevated PTH (secondary hyperparathyroidism) |
| Potassium | With every RAAS-blocking or nsMRA dose change | Hyperkalemia that threatens continuation of nephroprotective therapy |
| Fasting lipid panel | At CKD diagnosis | Statin eligibility by age/eGFR/risk category (not a treat-to-target recheck) |